Development of translational models and identification of novel therapeutic targets for medulloblastoma

Medulloblastoma (MB), the most frequent malignant brain cancer in children, has traditionally been studied using established cell lines, patient-derived xenografts, engineered mouse models, and orthotopic transplants. While these models have been instrumental in advancing our understanding of MB pathogenesis, they predominantly represent later stages of tumor development, failing to adequately capture the early events of tumor initiation and progression. Moreover, many preclinical in vivo models are unable to fully recapitulate human tumor pathogenesis due to species differences in development and organ structures. Additionally, despite the discovery of numerous molecular therapeutic strategies over recent decades, only a small fraction of drugs demonstrating preclinical efficacy have translated successfully into human clinical trials. This further underscores the need for innovative approaches to model tumorigenesis. Modeling early tumor development has become possible with the emergence of induced pluripotent stem cell (iPSC) technology, thereby opening new avenues for the discovery of potential therapeutic targets. This thesis describes the development of novel, disease-relevant model systems that integrate patient- derived iPSCs and zebrafish xenografts to faithfully recapitulate the early stages of MB tumorigenesis. The overall aim of this thesis was to develop both in vitro and in vivo models that mimic MB initiation and progression, thereby facilitating mechanistic studies and the identification of novel therapeutic targets.

In Paper I, we established a new human model system for sonic hedgehog (SHH)- driven MB by utilizing neuroepithelial stem (NES) cells derived from non- cancerous Gorlin syndrome patient iPSCs. Re-transplantation of these cells leads to a more malignant phenotype and accelerated tumor formation, providing a robust platform for studying the molecular changes during tumor development. Importantly, this study identified LGALS1 (galectin-1) as a direct GLI target gene that is upregulated in SHH MB, recognizing it as a promising novel therapeutic target. Furthermore, we describe the development of an orthotopic zebrafish- based xenograft model for MB in Paper II. In this model, MB cells exhibited intrinsic homing to the hindbrain region of developing zebrafish embryos upon blastula- stage transplantation, a process further enhanced by in vitro neural stem cell culture conditions. The blastula-stage zebrafish xenografts not only recapitulated tumor growth and neurotropism across MB subgroups but also serve as an efficient in vivo system for rapid drug testing. This provides a scalable alternative to traditional mouse models. Both of these novel model systems have been used to elucidate the role of galectins in SHH-driven MB (Paper III). This study demonstrated that LGALS3 (galectin-3) is highly expressed in these tumors and plays a crucial role in maintaining primary cilia structures, thereby facilitating canonical SHH signaling. Functional studies revealed that loss or inhibition of galectin-3 reduces tumorigenicity by impairing cell proliferation, migration and overall tumor progression in vivo. Additionally, galectin-3 modulates amino acid uptake, particularly arginine, through a TGF-B-mediated transcriptional mechanism on the activity of its solute carrier (SLC) transporters. All of this highlights its potential as a therapeutic target for SHH-driven MB. Lastly, we present the development of a human in vitro iPSC-derived cerebellar organoid model to explore their potential as a model system for MB (Paper IV). These organoids, generated from both control and Gorlin syndrome patient-derived iPSCs, recapitulated critical aspects of cerebellar development. Notably, Gorlin- derived organoids displayed accelerated growth, increased proliferation, and a higher abundance of PAX6-positive cells compared to controls. This makes them an invaluable, all-human system for further exploration of MB pathogenesis.

In conclusion, this thesis significantly advances MB research by developing innovative in vitro and in vivo models that more accurately mirror the early stages of MB tumor initiation and progression. Additionally, we demonstrate that these model systems enable the discovery of novel therapeutic targets for SHH-driven MB, holding promise for translational research.

List of scientific papers

I. Susanto E, Marin Navarro A, Zhou L, Sundström A, van Bree N, Stantic M, Moslem M, Tailor J, Rietdijk J, Zubillaga V, Hübner JM, Weishaupt H, Wolfsberger J, Alafuzoff I, Nordgren A, Magnaldo T, Siesjö P, Johnsen JI, Kool M, Tammimies K, Darabi A, Swartling FJ, Falk A, Wilhelm M. Modeling SHH-driven medulloblastoma with patient iPS cell-derived neural stem cells. Proc Natl Acad Sci U S A. 2020 Aug 18; 117(33) : 20127-20138. https://doi.org/10.1073/pnas.1920521117

II. van Bree N, Oppelt AS, Lindstrom S, Zhou L, Boutin L, Coyle B, Swartling FJ, Johnsen JI, Brautigam L, Wilhelm M. Development of an orthotopic medulloblastoma zebrafish model for rapid drug testing. Neuro Oncol. 2025 Mar; 27(3) : 779-794. https://doi.org/10.1093/neuonc/noae210

III. van Bree N, Oppelt AS, Zimmer E, Wiesinger ML, Kvastad L, Escudero Morlanes J, Marques R, Bell N, Boutin L, Dias J, Zetterberg F, Nilsson U, Schäfer S, Zhou L, Teixeira A, Wilhelm M. Galectin-3 supports tumorigenesis in SHH-driven medulloblastoma by coordinating ciliary signaling and amino acid transport. [Manuscript]

IV. van Bree N, Haazen L, Falk A, Magnaldo T, Wilhelm M. Generation of Gorlin syndrome patient iPSC-derived cerebellum organoids as a model system for SHH-driven medulloblastoma. [Manuscript]